KR102322709B1 - Image processing method, image processing circuit and display device using the same - Google Patents

Abstract

The present invention relates to an image processing method and circuit capable of minimizing image quality degradation of an HDR image to be displayed on an SDR display device without gamma conversion of a data drive IC, and a display device using the same.
In an image processing method of the present invention, in order to display a first image having HDR on a display device having SDR, a gamma curve having a cumulative minimum luminance error with the first image among a plurality of gamma curves corresponding to the display device is generated. selecting; and converting the first image into a second image having an SDR according to the selected gamma curve.

Description

Image processing method, image processing circuit, and display device using the same

The present invention relates to a display device, and in particular, an image processing method and image processing that can be displayed on a Standard Dynamic Range (SDR) display device by minimizing the quality degradation of a High Dynamic Range (HDR) image It relates to a circuit and a display device using the same.

In general, in order to display a photographed image on a display device, digitization of the image is required. In this case, gamma encoding and gamma decoding are required. Gamma encoding aims to contain the maximum amount of information within a given bandwidth (eg, 256 gray levels for an 8-bit video signal), and is relatively sensitive to brightness changes in the low luminance section compared to the high luminance section due to the characteristics of human visual perception, i.e., a non-linear characteristic. has In consideration of this, a non-linear transfer function is used for gamma encoding, which is defined in the Rec. (Recommendation) 709 and Rec. 1886 standards with the reciprocal of 2.4 as the exponent. The display device determines the gamma reference voltage in consideration of a function having an exponent of 2.4, which is an inverse function of a transfer function used during encoding, in order to convert a gamma-encoded image into an originally intended luminance for each gray level.

The conventional Rec.709 standard time display device is a CRT (Cathode-Ray Tube), so it has a narrow dynamic range of ^{about 0 to 100 cd/m 2 .} It does not fit the visual perception characteristics. Real humans have a wide dynamic range of ^{approximately 10 -4} to 10 ⁸ cd/m ^{2 in the real world.} A technology that takes this into consideration is High Dynamic Range (HDR), and until now, HDR technology has been mostly focused on the camera field.

Recently, there has been a movement to expand HDR to video production and display development, and representatively, SMPTE (Society of Motion Picture and Television Engineers) ST. (Standard) 2084 standard, BDA (Bluray Disc Association) HDR standard, etc. were enacted. or are under discussion. The SMPTE ST.2084 standard refers to an Electro-Optical Transfer Function (EOTF) that encodes an HDR image for an HDR display device, and is also called Perceptual Quantzer (PQ).

As described above, gamma encoding aims to contain the maximum amount of information within a given bandwidth, and decoding is a process for converting encoded information into an original brightness representation. Accordingly, since encoding and decoding are inversely functional, image quality deterioration cannot be avoided if the functions of encoding and decoding are different.

That is, the HDR image should have better quality than the SDR image, but when the HDR image is displayed on a conventional standard dynamic range (SDR) display device, the encoding and decoding functions are different and the image quality is deteriorated compared to the SDR image. There is a problem.

This is because the HDR image encoded with the HDR standard (ST.2084) was not properly decoded as most conventional SDR display devices decode using the gamma defined by the conventional SDR standard (Rec.709/Rec.1886). , this problem is not solved no matter how wide the dynamic range of the display device is.

Contrary to the above case, if there is a display device conforming to the HDR standard (ST.2084), the HDR image will be accurately expressed, but the SDR image will not be accurately expressed.

In order to solve these problems, a decoding function that exactly corresponds to an image-encoded transfer function must be implemented in the display device. Therefore, in order to properly represent both SDR and HDR images from the viewpoint of a display device, it is ideal to implement each decoding function (EOTF) of SDR and HDR in the data drive IC, but there is a problem in that the cost increases.

The present invention has been devised to solve the above-mentioned problems of the prior art, and the problem to be solved by the present invention is an image processing method capable of displaying on an SDR display device by minimizing the deterioration of the image quality of an HDR image without gamma conversion of a data drive IC. and a circuit and a display device using the same.

In order to solve the above problems, in an image processing method according to an embodiment of the present invention, in order to display a first image having HDR on a display device having SDR, the first image among a plurality of gamma curves corresponding to the display device is displayed. selecting a gamma curve having an image and a cumulative minimum luminance error; and converting the first image into a second image having an SDR according to the selected gamma curve.

The step of selecting the gamma curve having the cumulative minimum luminance error comprises calculating the luminance error for each frame by mapping the first image to each of the plurality of gamma curves having different maximum luminance, and accumulating the calculated luminance error; detecting a cumulative luminance error for each of the plurality of gamma curves; selecting a gamma curve having the cumulative minimum error from among the cumulative luminance errors for each of the plurality of gamma curves; It includes the step of determining and outputting.

The converting of the first image into the second image includes selecting a LUT corresponding to the determined maximum luminance from a lookup table for HDR-to-SDR conversion (hereinafter referred to as LUT) preset corresponding to each of the plurality of gamma curves. and mapping the first image to the second image using the selected LUT.

In the image processing method of the present invention, before the step of selecting any one gamma curve, determining a roll-off inflection point according to image characteristics according to the analysis result of the first image; The method further includes the step of roll-off processing a high gradation above the determined roll-off inflection point.

In the determining of the roll-off inflection point, the high grayscale frequency of the top n% (a natural number less than 100) is calculated by analyzing the histogram of the first image, and the roll-off inflection point is determined according to the calculated high grayscale frequency. adaptively determined.

The image processing method of the present invention includes the steps of determining whether an input image is an HDR image or an SDR image according to option information before analyzing the first image; and if the input image is the SDR image, the input image is converted passing, and supplying the input image as the first image when the input image is the HDR image.

In order to display a first image having HDR on a display device having SDR, the image processing circuit according to an embodiment of the present invention determines a roll-off inflection point according to image characteristics according to a result of analyzing the first image, and , a roll-off processing unit for roll-off processing a high gradation equal to or greater than the determined roll-off inflection point in the first image; selecting a gamma curve having a cumulative minimum luminance error with the first image from among a plurality of gamma curves corresponding to the display device, and converting the first image into a second image having the SDR according to the selected gamma curve; It includes an image mapping unit.

The roll-off processing unit analyzes the histogram of the first image and calculates and outputs a high grayscale frequency of higher than n% (a natural number less than 100) and outputs the histogram analysis unit, and roll-off according to the calculated high grayscale frequency and a roll-off inflection point determiner that adaptively determines an inflection point, and a roll-off operator that rolls-off a high grayscale greater than or equal to the determined roll-off inflection point in the first image.

The image mapping unit maps the HDR image to each of the plurality of gamma curves having different maximum luminance, calculates a luminance error for each frame, accumulates the calculated luminance error, and accumulates the luminance error for each of the plurality of gamma curves a cumulative luminance error detector for detecting , selects a LUT corresponding to the determined maximum luminance from among LUTs for HDR-to-SDR conversion preset corresponding to each of the plurality of gamma curves, and converts the first image into the second image using the selected LUT It includes an HDR-to-SDR converter for mapping.

The image processing circuit of the present invention is located in front of the roll-off processing unit, determines whether an input image is an HDR image or an SDR image according to option information, and bypasses the input image if the input image is the SDR image , when the input image is the HDR image, further comprising a content selection unit for supplying the input image as the first image to the roll-off processing unit.

A display device according to an exemplary embodiment includes a display panel, the above-described image processing circuit, a panel driver for displaying an image supplied from the image processing circuit on the display panel, and a timing for controlling a driving timing of the panel driver Includes controller. The image processing circuit is embedded in the timing controller, located between the timing controller and the panel driver, or located in front of the timing controller.

The display device of the present invention includes a backlight unit irradiating light to the display panel, and adjusting the luminance of the backlight unit in response to a dimming value output from the timing controller using the maximum luminance determined by the image processing circuit. It additionally includes a backlight driver.

The image processing method and circuit according to the present invention and a display device using the same perform data conversion on an HDR image encoded with the HDR standard in the SDR display device like an image encoded with the SDR standard, so that the SDR image and HDR image without gamma conversion of the data drive IC All images can be implemented with image quality degradation with minimum error.

In other words, the image processing method and circuit according to the present invention and a display device using the same minimize high grayscale saturation by adaptively determining a roll-off inflection point through analysis of an HDR image, and the cumulative minimum for an HDR image. By determining a gamma curve having a luminance error and converting an HDR image into an SDR image, grayscale loss due to HDR-to-SDR mapping can be minimized. Accordingly, it is possible to output the HDR image from the SDE display device by minimizing the deterioration of the image quality of the HDR image.

1 is a graph illustrating a comparison between a PQ encoding curve according to an HDR transfer function (ST.2084) and a gamma encoding curve according to an SDR transfer function (Rec.1886) in order to facilitate understanding of the present invention.
2 is a graph illustrating a comparison between a PQ decoding curve conforming to the HDR transfer function (ST.2084) and a gamma decoding curve conforming to the SDR transfer function (Rec.1886) in order to help the understanding of the present invention.
3 is a diagram illustrating a case in which grayscale loss occurs when a grayscale value of an HDR image is mapped along a 2.2 gamma curve in order to facilitate understanding of the present invention.
4 is a block diagram schematically illustrating the configuration of an image processing circuit according to an embodiment of the present invention.
FIG. 5 is a graph illustrating a roll-off process for minimizing high grayscale saturation in the roll-off processor shown in FIG. 4 compared with a clipping process.
FIG. 6 is a view for explaining a method of adaptively determining a roll-off start point according to image characteristics in the roll-off processing unit shown in FIG. 4 .
7 is a block diagram illustrating an internal configuration of the cumulative luminance error calculator shown in FIG. 4 .
8 is a flowchart illustrating an image processing method step-by-step according to an embodiment of the present invention.
9 is a block diagram schematically illustrating a liquid crystal display to which an image processing circuit is applied according to an exemplary embodiment of the present invention.

1 is a graph illustrating a comparison between a PQ encoding curve according to an HDR transfer function (ST.2084) and a gamma encoding curve according to an SDR transfer function (Rec.1886) in order to facilitate understanding of the present invention.

SMPTE ST.2084 considering HDR display has a dynamic range of 0 to 10,000 cd/m ² , which is determined in consideration of a much wider dynamic range than 0 to 100 cd/m ^{2 of the conventional SDR display.} Accordingly, there is a big difference between the ST.2084 transfer function for encoding an HDR image and the BT.1886 transfer function for encoding an SDR image.

Referring to FIG. 1, a PQ encoding curve (red) showing a grayscale versus luminance graph according to the HDR standard ST.2084 transfer function, and a gamma encoding curve showing a grayscale versus luminance graph according to the SDR standard BT.1886 transfer function ( Blue) shows a very large difference.

The present invention proposes a method for implementing an SDR image and an HDR image with minimal image quality degradation without gamma conversion of a drive IC in an SDR display device conforming to the SDR transfer function (Rec.709 / Rec.1886).

For this purpose, the basic concept of the present invention is to convert an image encoded with the HDR transfer function (ST.2084), that is, the PQ encoded image, like an image encoded with the SDR transfer function (Rec.709 / Rec.1886).

FIG. 2 shows a comparison of the PQ curve at the decoding time according to the HDR transfer function ST.2084 (= PQ) and the 2.2 gamma curve at the decoding time according to the SDR transfer function in order to help the understanding of the basic concept of the present invention.

Referring to FIG. 2 , the 150 gradation value of the PQ-encoded image should be expressed as approximately 250 nit. in the display device, but in the SDR display device with a maximum of 400 nit. . For this reason, as described in the prior art, the HDR image is displayed darkly in the SDR display device.

On the other hand, referring to FIG. 2 , when the 150 gray level of the HDR image is changed (mapped) to 180 gray level in the SDR display device, the same 250 nit.

However, such a simple data mapping method may cause several problems. First, since 1:1 matching is impossible in the same bandwidth (for example, assuming that both PQ and 2.2 gamma are 8 bits), grayscale loss is inevitable. Second, since the dynamic range of PQ is generally wider than that of the display device, 200 gradations of the PQ-encoded image in FIG. 2 are expected to exceed about 1,000 nits, but it is impossible to express in SDR display devices of up to 400 nits. The luminance of grayscales is saturated, and grayscale clustering occurs at high grayscales.

FIG. 3 is a diagram illustrating a case in which grayscale loss occurs when data mapping of grayscale values of the above-described PQ-encoded HDR image along the 2.2 gamma curve occurs.

Referring to FIG. 3 , it can be seen that as the maximum luminance 400 , 800 , 1000 , 1500 , 2000 and 4000 of the SDR display increases, the loss at a low gray level may increase. In a fragmentary way, 16 grayscales encoded by PQ are mapped to 6 grayscales in a 2.2 gamma display device with a maximum of 400 nit.

In order to solve these problems, the present invention provides an image for minimizing grayscale loss due to mapping an HDR image according to the SDR gamma curve and high grayscale saturation that may occur when the dynamic range of the HDR image is wider than the dynamic range of the SDR display device. A processing method and an image processing circuit are proposed.

4 is a block diagram schematically illustrating the configuration of an image processing circuit according to an embodiment of the present invention.

The image processing circuit 50 illustrated in FIG. 4 includes a content selection unit 10 , a roll-off processing unit 20 , and an image mapping unit 30 . The roll-off processing unit 20 includes a histogram analysis unit 22 , a roll-off inflection point determination unit 24 , and a roll-off calculation unit 26 . The image mapping unit 30 includes a cumulative luminance error detector 32 , a maximum luminance determiner 34 , and an HDR-to-SDR converter 36 .

The content selection unit 10 receives an input image RGB and option information from the outside, and determines whether the input image RGB is an HDR image or an SDR image according to the option information. The option information includes image information indicating whether the input image RGB is an HDR image or an SDR image. If the input image RGB is an HDR image, the content selector 10 outputs it to the roll-off processing unit 20 , and if the input image RGB is an SDR image, it outputs it to the data driver.

Before mapping the HDR image supplied from the content selection unit 10 to the SDR image, the roll-off processing unit 20 adjusts the overall luminance of the high grayscale region to be darkened in order to minimize the high grayscale saturation due to the mapping. A roll-off processing technique is used. In particular, the roll-off processing unit 20 adaptively performs the roll-off processing according to the image characteristics analyzed by the input HDR image. The roll-off processor 20 adaptively determines an inflection point (roll-off starting point) indicating a grayscale position at which roll-off starts according to image characteristics, thereby minimizing high grayscale saturation and image quality deterioration due to the inflection point. In other words, the roll-off processing unit 20 adaptively determines the roll-off inflection point (roll-off starting point) according to the high grayscale frequency of the upper n% or more by analyzing the HDR image-based histogram, and For gradation, roll-off processing is performed and output.

To this end, the roll-off processing unit 20 analyzes the DR image-based histogram and outputs the high grayscale frequency of the top n% or more, and the histogram analyzer 22 outputs the high grayscale frequency from the histogram analyzer 22. A roll-off inflection point determining unit 24 for adaptively determining a roll-off inflection point, and a roll-off calculating unit 26 for performing a roll-off process for high grayscales greater than or equal to the determined inflection point.

5 is a graph for explaining a roll-off method of the roll-off processing unit 20 shown in FIG. 4 .

In FIG. 5 , a blue dotted line indicates a simple clipping method when the dynamic range of the input image is greater than the dynamic range of the display device. For example, if the luminance corresponding to the 170 gradation in the HDR image is 400 nit., all high gradations higher than 170 gradations are changed to 170.

On the other hand, the roll-off method is a method in which an arbitrary gray scale is determined as an inflection point as shown in the red dotted line shown in FIG. 5 and bent. This can reduce high grayscale saturation compared to clipping, but since image quality deterioration can be recognized based on the inflection point, the inflection point is adaptively determined through analysis of the input image. In other words, the roll-off processor 20 adaptively determines an inflection point according to image characteristics through a histogram analysis of the HDR image, thereby minimizing high grayscale saturation and image quality deterioration due to the inflection point.

FIG. 6 is a diagram for explaining a method of adaptively determining a roll-off start point according to image characteristics in the roll-off processing unit 20 illustrated in FIG. 4 .

6(a) and 6(b) are examples of histogram analysis of HDR images, wherein the X-axis represents normalized luminance and the Y-axis represents the frequency.

The transfer function (ST.2084) for encoding the HDR image, that is, the PQ EOTF is as defined in Equation 1 below, and the luminance for the input grayscale can be obtained using Equation 1 below.

In Equation 1, L is luminance, N is input grayscale, and m ₁ to m ₃ , and c ₁ to c ₃ are constant values. For example, m ₁ = 2610/4096×(1/4) = 0.1593017578125, m ₂ = 2523/4096×128 = 78.84375, c ₁ = 3424/4096 = 0.8359375 = c ₃ -c ₂ +1, c ₂ = 2413 /4096×32 = 18.8515625, c ₃ = 2392/4096×32 = 18.6875.

In the case of a dark image with a small high grayscale region as shown in Fig. 6(a), the roll-off inflection point may be located at a higher grayscale level, but in the case of a bright image with many high grayscale regions as shown in Fig. 6(b), the roll-off inflection point is the roll-off inflection point. It is better to go toward the lower gradation level, and it is decided by considering the brightness aspect of the image.

_{Specifically, the roll-off inflection point determiner 24 determines the roll-off inflection point (Roll-off pos} ) as shown in Equation 2 below in consideration of the frequency of the upper n% high grayscale region from the histogram analysis unit 22 . decide

<Equation 2>

If NumberOfGray(n) > Thereshold = Roll-off _pos = (1 - a ) × Roll-off _initial

Else Roll-off _pos = Roll-off _initial

If the frequency NumberOfGray(n) of the upper n% high grayscale region is greater than the threshold, the roll-off inflection point (Roll-off _pos ) is determined as "(1 - a)×Roll-off _initial ", and the other rolls The -off inflection point (Roll-off _pos ) is determined by the initially set roll-off initial inflection point (Roll-off _initial). Here, a is an experimental constant value, which increases as the luminance increases and decreases as the luminance decreases. a is an empirical balance by an experiment, and the minimum and maximum values are predetermined and may be set to linearly proportional values. Roll-off initial inflection point (Roll-off _initial) is previously set according to the maximum luminance of the display device.

_{The roll-off operation unit 26 multiplies the roll-off inflection point (Roll-off pos} ) determined according to the above-described image analysis by the roll-off inflection point determiner 24 and the input grayscale (Gray _in ) as shown in Equation 3 below. By calculating, the roll-off-processed output gradation (Gray _out ) is output.

<Equation 3>

Gray _out = (Roll-off _pos )×(Gray _in )

For example, a maximum value (GrayMax) among R, G, and B grayscale data (0≤GrayMax≤255 when grayscale data is 8 bits) is detected for each pixel in an image having a size of 100×100 pixels. When a histogram is created with the maximum value (GrayMax) for each pixel, the X-axis becomes a grayscale value in the range of 0 to 255, and the Y-axis becomes the frequency. _{For example, the roll-off inflection point (Roll-off pos} ) is determined in consideration of the frequency (100×100×0.1) of the upper 10% high grayscale region.

The grayscale value (Xm) that satisfies the condition of Initial X=255, (Histogram[X] + Histogram[X-1] + ... + Histogram[Xm]) > (100×100×0.1) is the uppermost It corresponds to the 10% high gradation area. If the overall image is dark, (X-m) will be close to 0, and if it is a bright image, it will be close to 255.

For example, assuming that the threshold for determining the roll-off inflection point is 192 gradations, if (Xm) is less than 192, a=0, and it is determined that the roll _{-off inflection point (Roll-off pos) adjustment is not necessary.} It is determined by the roll-off _initial. On the other hand, when (Xm) is less than 192, a>0 becomes a roll-off inflection point (Roll-off _pos ) is changed to a side closer to 0 than the _{roll-off initial} inflection point (Roll-off initial).

4 , in order to minimize grayscale loss due to mapping, the image mapping unit 30 selects a gamma curve that minimizes image quality degradation, and maps the HDR image to the SDR image using the selected gamma curve. In other words, the image mapping unit 30 receives the roll-off-processed HDR image through the roll-off processing unit 20 , calculates a cumulative luminance error with a plurality of gamma curves having different maximum luminances, and calculates the minimum Image mapping is performed using a gamma curve having a cumulative luminance error.

To this end, the image mapping unit 30 maps the HDR image input from the roll-off processing unit 20 along a plurality of gamma curves having different maximum luminances, respectively, to calculate and accumulate the luminance error for each frame, thereby accumulating the luminance error. a cumulative luminance error detection unit 32 for detecting 34) and HDR-to-SDR that converts an HDR image to an SDR image using an HDR-to-SDR LUT (Look-up Table) according to the determined maximum luminance (L) and outputs the converted SDR image to the data driver. A conversion unit (36) is provided. Here, the cumulative luminance error detector 32 and the HDR-to-SDR converter 36 may be implemented in the form of LUTs.

A method of minimizing image quality degradation due to mapping in the image mapping unit 30 will be described later.

As can be seen from FIG. 2 described above, as the maximum luminance of the gamma curve increases, low grayscale mapping becomes difficult and high grayscale mapping becomes easy. With these characteristics and in HDR-to-SDR mapping, it can be seen that the image quality deterioration is minimized as all pixels in the PQ-encoded image express the original luminance as much as possible. Select and map the minimized gamma curve.

In other words, the image mapping unit 30 is 100 ~ Max nit. A gamma curve having a minimum cumulative luminance error is selected by calculating a cumulative luminance error with respect to the gamma curves, and image mapping is performed with the selected gamma curve.

When the cumulative luminance error is calculated through an equation, the circuit load may be increased, so the accumulated luminance error detection unit 32 is implemented in the form of a LUT in advance by using Equation 4 below.

<Equation 4>

i = input grayscale, n is the number of bits, r is gamma exp (eg 2.2),

PQ(i) see Equation 1, Gamma(i) = (i / (2 ⁿ - 1)) ^r ,

Minimum Luminance Difference LUT(i) = (PQ(i)- Gamma(i))/PQ(i)

The maximum luminance determiner 34 selects a gamma curve having the minimum accumulated luminance error from among the accumulated luminance errors output from the accumulated luminance error detection unit 32, and sets the maximum luminance L of the selected gamma curve as HDR-to-SDR. output to the converter 36 . Meanwhile, in order to prevent flickering due to sudden change or noise, the maximum luminance determiner 34 sets the maximum luminance L output from the maximum luminance determiner 34 to adjacent frames to which weights are assigned using a time filter. It is possible to determine the maximum luminance (L) by averaging during the period. As the time filter, an IIR (Infinite Impulse Response) filter can be used.

When the maximum luminance L is determined by the gamma curve having the minimum accumulated luminance error in the maximum luminance determiner 34, the HDR-to-SDR converter 36 performs an HDR-to- Convert HDR image to SDR image using SDR LUT. The HDR-to-SDR LUT is implemented in advance through Equation 5 below. PQ(i) is according to Equation 1 mentioned above and has a value in the range of 0 to 1. 0 to n grayscales (n is determined according to the maximum number of bits of the display device. 8-bit is 255, r=gamma of the display device) can be defined as in Equation 5 below.

<Equation 5>

HDR-to-SDR LUT(i) = Power(PQ(i) * 10,000 / L, 1 / r) × n

The HDR-to-SDR converter 36 includes a plurality of LUTs according to the maximum luminance L of the gamma curve, and selects the LUT corresponding to the maximum luminance L determined by the maximum luminance determiner 34, An HDR image can be converted to an SDR image through the selected LUT. At this time, the HDR-to-SDR converter 36 supplies the determined maximum luminance L and the maximum luminance L of the gamma curve having the minimum cumulative luminance error to the dimming controller (not shown), so that the dimming controller is configured to display the liquid crystal display. The maximum luminance (L) described above may be utilized to determine the dimming gain that controls the backlight luminance of the device.

Table 1 below shows, as an example, the result of calculating the minimum cumulative luminance error for seven images by the cumulative luminance error detection unit 32 shown in FIG. 4 .

In Table 1, the left luminance corresponding to the red number corresponds to the maximum luminance L of the gamma curve having the minimum cumulative luminance error.

Referring to Table 1, images #1, #2, and #4 have the minimum cumulative luminance error in the 100 nit gamma curve, #3 images in the 200 nit gamma curve, #5, #7 images in the 300 nit gamma curve, #6 It can be seen that the image has a minimum cumulative luminance error in the 500 nit gamma curve. Accordingly, the maximum luminance L of the gamma curve having the minimum cumulative luminance error may be determined according to the characteristics of the image.

8 is a flowchart illustrating an image processing method according to an embodiment of the present invention step by step, which is performed by the image processing circuit 50 shown in FIG. 4 , and thus will be described in conjunction with FIG. 4 .

When an image RGB is input to the content selection unit 10 shown in FIG. 4 in step 2 (S2), it is determined whether the input image RGB is an HDR image or an SDR image using the option information in step 4 (S4). judge

If the input image RGB is determined to be an HDR image in step 4 (S4), the roll-off processing unit 20 performs high grayscale saturation by data mapping before mapping the HDR image in step 6 (S6). In order to minimize the input HDR image, roll-off processing is adaptively performed according to the analyzed image characteristics. The roll-off processing unit 20 adaptively determines a roll-off inflection point (roll-off starting point) according to the high grayscale frequency of the top n% or more by analyzing the histogram based on the HDR image, and rolls for the high grayscale greater than or equal to the determined inflection point. -Off process and output.

In step 8 (S8), the image mapping unit 30 selects a gamma curve that minimizes image quality degradation in order to minimize grayscale loss due to image mapping, and converts the HDR image using the gamma curve selected in step 10 (S10). Mapping to SDR image. The image mapping unit 30 calculates the accumulated luminance error for the HDR image supplied from the roll-off processing unit 20 for each gamma curve by using the accumulated luminance error LUT respectively set according to a plurality of gamma curves having different maximum luminance. is detected, a gamma curve having the smallest cumulative luminance error is selected, and an HDR image is converted into an SDR image using the selected gamma curve.

In step 12 ( S12 ), the SDR image converted in step 10 ( S10 ) or the SDR image determined in step 4 ( S4 ) is output to the data driver.

When the HDR image is displayed on the SDR display device by applying the image processing method according to the embodiment of the present invention, the luminance can be increased and the image quality can be improved, compared to the case of simple data mapping of the HDR original image.

Accordingly, in the present invention, even in an SDR display device that does not comply with the HDR standard (ST.2084), by reproducing the HDR image encoded with ST.2084 by minimizing the image quality degradation, the SDR image and the HDR image can be selectively selected without changing the gamma of the data driver. Since it can display, it is possible to minimize the cost increase of data drivers and timing controllers.

The image processing circuit and method of the present invention described above may be applied to liquid crystal display devices, organic light emitting diode displays, and the like.

9 is a block diagram illustrating an example of a liquid crystal display to which an image processing circuit according to an embodiment of the present invention is applied.

The liquid crystal display shown in FIG. 9 includes a timing controller 100 , a data driver 200 and a gate driver 300 serving as a panel driver, a display panel 400 , a gamma voltage generator 500 , a backlight unit 600 , It includes a backlight driver 700 , a power supply not shown, and the like.

The display panel 400 displays an image through a pixel array in which pixels are arranged in a matrix form. Each pixel of the pixel array is composed of red (R; hereinafter R), green (Green; hereinafter G), and blue (Blue; hereinafter B) sub-pixels. Alternatively, the R/W/B/G sub-pixels to which a white (W) sub-pixel having higher luminous efficiency than the RGB sub-pixel is added may be configured. Alternatively, each pixel can be As the display panel 50 , an LCD panel or an OLED panel may be applied.

The data driver 200 receives a data control signal and image data from the timing controller 100 . The data driver 200 is driven according to the data control signal, subdivides the reference gamma voltage set supplied from the gamma voltage generator 500 into gray voltages corresponding to gray values of data, and then divides the subdivided gray voltages. to convert digital image data into analog image data signals.

The data driver 200 includes a plurality of data drive ICs that divide and drive data lines of the display panel 400 , and each data drive IC includes a Tape Carrier Package (TCP), Chip On Film (COF), and Flexible Print (FPC). Circuit), etc. may be mounted on the display panel 400 and attached to the display panel 400 by a tape automatic bonding (TAB) method or may be mounted on the display panel 400 by a COG (chip on glass) method.

The gate driver 300 drives each of the plurality of gate lines of the display panel 400 using a gate control signal supplied from the timing controller 100 . The gate driver 300 supplies a scan pulse of a gate-on voltage to each gate line in a corresponding scan period in response to a gate control signal, and supplies a gate-off voltage in the remaining period. The gate driver 300 may receive a gate control signal from the timing controller 100 or may receive a gate control signal from the timing controller 100 via the data driver 200 . The gate driver 300 is composed of at least one gate IC and is mounted on a circuit film such as TCP, COF, FPC, etc. and attached to the display panel 400 in a TAB method, or mounted on the display panel 400 in a COG method. can In contrast, the gate driver 300 is formed on the thin film transistor substrate together with the thin film transistor array constituting the pixel array of the display panel 400 , so that the gate driver 300 is a gate in panel (GIP) type embedded in the non-display area of the display panel 400 . can be provided as

The timing controller 100 receives image data and timing signals from an external host system. The timing controller 100 outputs the input image data to the data driver 200 by performing image processing such as necessary image quality compensation. The timing controller 100 generates a data control signal and a gate control signal for controlling driving timings of the data driver 200 and the gate driver 300, respectively, using the input timing signals to generate the data driver 200 and the gate driver ( 300) respectively. The timing signal supplied from the host system to the timing controller 100 includes a dot clock, a data enable signal, a vertical synchronization signal, and a horizontal synchronization signal, but the vertical synchronization signal and the horizontal synchronization signal may be omitted here. When the vertical synchronization signal and the horizontal synchronization signal are omitted, the timing controller 100 may generate and use the vertical synchronization signal and the horizontal synchronization signal by counting the data enable signal according to the dot clock. The data control signals supplied from the timing controller 100 to the data driver 200 include a source start pulse, a source sampling clock, a polarity control signal, and a source output enable signal. The gate control signals supplied from the timing controller 100 to the gate driver 300 include a gate start pulse, a gate shift clock, and a gate output enable signal.

As shown in FIG. 9 , the image processing circuit 50 described above in FIG. 4 is embedded in the timing controller 100 , located between the timing controller 100 and the data driver 200 , or the timing controller 100 . ) can be located at the input terminal. The image processing circuit 50 determines whether the input image data is an HDR image or an SDR image, bypasses the SDR image, and maps the HDR image to an SDR image by minimizing high grayscale saturation and image quality degradation as described above. do. The image processing circuit 50 may minimize high gray level saturation by determining a roll-off inflection point according to the HDR image characteristics and performing roll-off processing of a high gray level equal to or greater than the inflection point. Also, the image processing circuit 50 may determine a gamma curve having an accumulated minimum luminance error for the HDR image and map the HDR image to the SDR image according to the determined gamma curve LUT, thereby minimizing image quality degradation.

Meanwhile, the image processing circuit 50 supplies the maximum luminance L determined from the gamma curve having the accumulated minimum luminance error to the dimming controller (not shown) built in the timing controller 100 . Accordingly, the dimming controller may determine a dimming value for controlling the luminance of the backlight unit 600 using the maximum luminance L determined by the image processing circuit 50 and supply it to the backlight driver 700 .

The backlight unit 600 uses a fluorescent lamp such as CCFL or EEFL, or a direct type or edge type backlight including an LED as a light source. The direct backlight includes a light source disposed in the entire display area to face the rear surface of the display panel 400 , a light guide plate disposed on the light source, and a plurality of optical sheets, and the light emitted from the light source is liquid crystal through the plurality of optical sheets. The panel 40 is irradiated. The edge type backlight includes a light guide plate facing the rear surface of the display panel 400 , a light source disposed to face at least one edge of the light guide plate, and a plurality of optical sheets disposed on the light guide plate, wherein light emitted from the light source is It is converted into a surface light source through the light guide plate and is irradiated to the display panel 400 through a plurality of optical sheets.

The backlight driver 700 adjusts the luminance of the backlight 600 according to the dimming value from the timing controller 100 . The backlight driver 700 controls the brightness of the backlight 600 by generating a pulse width modulation (PWM) signal having a duty ratio corresponding to the dimming value and driving the backlight 600 .

As described above, the image processing method and circuit according to the present invention and a display device using the same perform data conversion on an HDR image encoded with the HDR standard in the SDR display device like an image encoded with the SDR standard, thereby providing an SDR image without gamma conversion of a drive IC. Both and HDR images can be implemented with image quality degradation with minimal error.

Those skilled in the art from the above description will be able to see that various changes and modifications can be made without departing from the technical spirit of the present invention. Accordingly, the technical scope of the present invention should not be limited to the contents described in the detailed description of the specification, but should be defined by the claims.

10: content selection unit 20: roll-off processing unit
22: histogram analysis unit 24: roll-off inflection point determining unit
26: roll-off operation unit 30: image mapping unit
32: cumulative luminance error detection unit 34: maximum luminance determining unit
36: HDR-to-SDR converter 50: image processing circuit
100: timing controller 200: data driver
300: gate driver 400: liquid crystal panel
500: gamma voltage generator 600: backlight unit
700: backlight driver

Claims (12)

In order to display a first image having a high dynamic range (hereinafter referred to as HDR) on a display apparatus having a standard dynamic range (hereinafter referred to as SDR), the first image and the cumulative minimum luminance among a plurality of gamma curves corresponding to the display apparatus selecting a gamma curve having an error;
converting the first image into a second image having SDR according to the selected gamma curve;
The step of selecting a gamma curve having the cumulative minimum luminance error includes:
calculating a luminance error for each frame by mapping the first image to each of the plurality of gamma curves having different maximum luminance, and accumulating the calculated luminance error to detect the accumulated luminance error for each of the plurality of gamma curves step and
selecting a gamma curve having the cumulative minimum luminance error from among the cumulative luminance errors for each of the plurality of gamma curves;
and determining and outputting a maximum luminance of the selected gamma curve. delete The method according to claim 1,
Converting the first image to the second image
selecting an LUT corresponding to the determined maximum luminance from among a lookup table for HDR-to-SDR conversion (hereinafter referred to as LUT) preset corresponding to each of the plurality of gamma curves;
and mapping the first image to the second image by using the selected LUT. 4. The method according to claim 1 or 3,
Before the step of selecting any one of the gamma curves,
determining a roll-off inflection point according to an image characteristic according to a result of analyzing the first image;
The image processing method further comprising the step of roll-off processing a high grayscale greater than the determined roll-off inflection point in the first image. 5. The method according to claim 4,
The step of determining the roll-off inflection point is
An image processing method of analyzing a histogram of the first image to calculate a high grayscale frequency of higher than n% (a natural number less than 100), and adaptively determining the roll-off inflection point according to the calculated high grayscale frequency. 6. The method of claim 5,
Before analyzing the first image,
determining whether the input image is an HDR image or an SDR image according to the option information;
When the input image is the SDR image, the input image is bypassed, and when the input image is the HDR image, the input image is supplied as the first image. In order to display the first image having HDR on a display device having SDR,
a roll-off processing unit that determines a roll-off inflection point according to an image characteristic according to a result of analyzing the first image, and roll-offs a high grayscale greater than or equal to the determined roll-off inflection point in the first image;
selecting a gamma curve having a cumulative minimum luminance error with the first image from among a plurality of gamma curves corresponding to the display device, and converting the first image into a second image having the SDR according to the selected gamma curve; An image processing circuit including an image mapping unit. 8. The method of claim 7,
The roll-off processing unit
a histogram analysis unit that analyzes the histogram for the first image and calculates and outputs the frequency of high grayscales greater than or equal to the top n% (a natural number less than 100);
a roll-off inflection point determining unit for adaptively determining a roll-off inflection point according to the calculated high grayscale frequency;
and a roll-off calculator configured to roll-off a high gray level equal to or greater than the determined roll-off inflection point in the first image. 8. The method of claim 7,
The image mapping unit
Accumulation of calculating the luminance error for each frame by mapping the HDR image to each of the plurality of gamma curves having different maximum luminance, accumulating the calculated luminance error, and detecting the accumulated luminance error for each of the plurality of gamma curves a luminance error detection unit;
a maximum luminance determiner for selecting a gamma curve having the cumulative minimum luminance error from among the accumulated luminance errors for each of the plurality of gamma curves, and determining and outputting a maximum luminance of the selected gamma curve;
A LUT corresponding to the determined maximum luminance is selected from among LUTs for HDR-to-SDR conversion preset corresponding to each of the plurality of gamma curves, and the first image is mapped to the second image by using the selected LUT An image processing circuit including an HDR-to-SDR converter. 10. The method of claim 9,
It is located in front of the roll-off processing unit, determines whether an input image is an HDR image or an SDR image according to option information, and if the input image is the SDR image, bypasses the input image, and the input image is the HDR image In the case of an image, the image processing circuit further comprising a content selection unit for supplying the input image as the first image to the roll-off processing unit. display panel;
The image processing circuit according to any one of claims 8 to 10;
a panel driver configured to display the image supplied from the image processing circuit on the display panel;
a timing controller for controlling the driving timing of the panel driver;
The image processing circuit is embedded in the timing controller, positioned between the timing controller and the panel driver, or positioned at a front end of the timing controller. 12. The method of claim 11,
a backlight unit irradiating light to the display panel;
and a backlight driver configured to adjust the luminance of the backlight unit in response to a dimming value output from the timing controller using the maximum luminance determined by the image processing circuit.

Images